Table 2. Regression results for dependent variables.

Multiple linear regression results for all dependent variables plus P values calculated through permutation tests, where applicable. b_1obs = the intervention coefficient from the linear models fitted to the observed network for each dependent variable (i.e., b₁ in the model described in Materials and Methods). Effect sizes are given in units of percent change and number of friends in the text.

Friendship network variable (y)	Multiple linear regression on observed data
	b1obs	SE	t	95% CI	P	Pperm
Time 1
Closeness, normalized	−0.001	0.003	t(209) = −0.36	(−0.008 to 0.005)	0.72	0.69
Betweenness, normalized	−0.001	0.001	t(209) = −1.09	(−0.002 to 0.001)	0.28	0.17
Total degree	0.47	0.49	t(209) = 0.97	(−0.49 to 1.43)	0.34	0.30
Outdegree	0.05	0.30	t(209) = 0.15	(−0.55 to 0.64)	0.88	0.88
Indegree	0.42	0.26	t(209) = 1.64	(−0.08 to 0.93)	0.10	0.08
Total tie strength	0.06	0.20	t(209) = 0.31	(−0.33 to 0.45)	0.76	0.75
Out-tie strength	0.05	0.23	t(209) = 0.23	(−0.40 to 0.50)	0.82	0.84
In-tie strength	−0.01	0.24	t(209) = −0.05	(−0.49 to 0.47)	0.96	0.95
Time 2
Closeness, normalized	0.009	0.003	t(208) = 2.86	(0.003 to 0.015)	0.005	0.007
Betweenness, normalized	0.001	0.001	t(208) = 1.50	(−0.0002 to 0.002)	0.14	0.10
Total degree	0.99	0.30	t(208) = 3.25	(0.39 to 1.59)	0.001	0.01
Outdegree	0.78	0.23	t(208) = 3.41	(0.33 to 1.23)	<0.001	0.007
Indegree	0.24	0.15	t(208) = 1.60	(−0.06 to 0.54)	0.11	0.23
Total tie strength	0.16	0.18	t(208) = 0.84	(−0.21 to 0.52)	0.40	0.43
Out-tie strength	0.43	0.21	t(208) = 2.02	(0.01 to 0.85)	0.04	0.08
In-tie strength	0.18	0.19	t(208) = 0.92	(−0.20 to 0.56)	0.36	0.45
Number of old friends	0.48	0.21	t(209) = 2.26	(0.06 to 0.90)	0.02	–
Number of new friends	0.33	0.18	t(209) = 1.83	(−0.03 to 0.68)	0.07	–
Proportion of old (versus new) friends*	−0.03	0.06	t(150) = −0.57	(−0.15 to 0.08)	0.57	–

*Among those who identified any friends in the course at the end of the semester (N = 167).